[Eeglablist] Why most of good 'brain' ICs are 'dipolar' with show 'red'-centerd scalp topos, although 2/3 of the cortex is in sulci?

[Makoto Miyakoshi]

Hi Eugen,

I'm not a super specialist on this topic, but let me share my experience
with you as a possible alternative explanation.
The pitfall may not be in excluding monopole and higher-order ones in the
multipole expansion from the picture. Basically I agree with Ramesh that in
the case of scalp EEG recording, the contribution of monopolar source must
be zero in theory. Again, I'm not a specialist here, but at least the
current consensus is like that.

Instead, I want to draw your attention to another more humanly problem.
A dipole sheet, which is defined as a massive number of dipoles aligned in
parallel in a 2-D plane, has counterintuitively different spatial
projection profile. I call it a transducer array effect. You don't need any
complicated phase modulation controlled over the distribution. Just a
simple array placements on a plane can let it project the electric fields
much further. You would think as if a monopole would be contributing on top
of a modeled dipole. See the following data.
https://sccn.ucsd.edu/wiki/Makoto's_preprocessing_pipeline#Physiologically_invalid_deep_dipoles.3F_.28For_130.2C000_page_views.2C_07.2F26.2F2021_Update.29
https://sccn.ucsd.edu/mediawiki/images/c/cb/SupplementaryFiguresForSimuUDL_BSCR80.pdf
I found this problem when I asked why my single dipole estimation almost
always (1 out of 4 estimates!) goes to subcortical locations, which makes
anatomically no sense. I summarized my opinion here
https://urldefense.com/v3/__https://onlinelibrary.wiley.com/action/downloadSupplement?doi=10.1111*2Fejn.15131&file=ejn15131-sup-0001-SupInfo.docx__;JQ!!Mih3wA!G4zHbSFIhBoPBMHAap-1JemDZ2t7VObfAPTWa1BFqhWZM92642zQ5lxjborLV-gWU3_MYZXf2aiArM7JlEqkiTWngxw$ 

See the section titled 'What does it mean to have dipole sources in ‘Upper
Basal’?. 4 <#_Toc45876597>'

>From the hindsight, I think it is very misleading, at least psychologically
(I'm a psychologist), to use a single dipole to model a dipole sheet with a
substantial area. It seems this 'pitfall' is in Riera et al. (2012),
otherwise how can you conclude that monopole contribution is present in the
scalp EEG recording? But again, I'm not a specialist here. I'm open to your
criticism.

Makoto

On Wed, Dec 20, 2023 at 8:03 AM Евгений Машеров <emasherov at yandex.ru> wrote:

> Hi Makoto
>
> Thank you very much for the interesting and valuable article Pitfalls in
> the dipolar model for the neocortical EEG sources: Riera JJ, Ogawa T, Goto
> T, Sumiyoshi A, Nonaka H, Evans A, Miyakawa H, Kawashima R,
> Doubts about the dipole model arose in me after my experience using the
> dipole localization program (BrainLoc by Yuri Koptelov). It used a
> single-dipole (it was also possible to require a search for two dipoles for
> each sample) model without forcing their coordinates with a selection of
> dipoles found throughout the EEG segment, leaving those points for which
> the single-dipole approximation of the signal is quite accurate, so it can
> be assumed that at the moment there is one dipole of high amplitude, and
> all the others have low amplitude and can be considered as noise. The found
> sources could be verified by an intracranial encephalogram during surgery,
> and the accuracy was quite high. However, in the case of deep sources, the
> signal attenuation from the dipole would be too high, and impossibly large
> amplitudes at the source would have to be assumed to obtain the signal
> observed on the scalp.
> A possible resolution of the contradiction would be to take into account
> the potential of not only dipole sources, but also monopoles.
> The specified dipole localization program, when trying to search for two
> dipoles simultaneously, usually did not give a good result, with the
> exception of searching for the source of a sharp wave of a peak-wave
> complex. In this case, two oppositely directed dipoles were observed
> emanating from the same point. It is possible that in reality there is a
> quadrupole source approximated by two dipoles.
> As for the connection between the activity of the thalamus and the
> occipital cortex, I can only assume that this may be a synchronization
> effect (like the synchronization of two pendulums connected by a weak
> spring). The synchronization mechanism can be through nerve impulses,
> purely electrically (ephaptically), or by diffusion of ions. But these are
> just assumptions that require criticism.
>
> Perhaps you can find a clearer explanation here
>
> https://urldefense.com/v3/__https://istina.msu.ru/download/621221220/1rFvVB:v1u5Qu1hRQWAPGg01ncDoTXd1eU/__;!!Mih3wA!HZ5DcvRqb9S9VAXTPN0K6cpV72eEGh00KN4xSwqmIfaX3t7othantbgfkgtqwwnc33qjGpEkTFWObqhaTCJVme8$
> PowerPoint presentation
>
> https://urldefense.com/v3/__https://istina.msu.ru/download/621282647/1rFw9L:1mcijNri7_vw2sz2onVQOFByo9A/__;!!Mih3wA!HZ5DcvRqb9S9VAXTPN0K6cpV72eEGh00KN4xSwqmIfaX3t7othantbgfkgtqwwnc33qjGpEkTFWObqhaw3IoP6I$
> And this is a small demo script in MATLAB.
>
> https://urldefense.com/v3/__https://istina.msu.ru/download/621281264/1rFvWt:v2A1PXPZdikElRoSyGK9sMlyBjs/__;!!Mih3wA!HZ5DcvRqb9S9VAXTPN0K6cpV72eEGh00KN4xSwqmIfaX3t7othantbgfkgtqwwnc33qjGpEkTFWObqhaUWc-C-E$
>
> Understanding that the material presented is very crude and not fully
> developed, I still hope that something useful will be found in it, and I
> will also be grateful for harsh criticism.
>
> Eugen Masherov
>
> > Hi Eugen,
> >
> > Thank you for your comment!
> >
> >> Therefore, the amplitude of the (summarized) recorded activity depends
> on
> >
> > the distance between the place of generation and the point of activity
> > registration. This may be due to the difference in the frequency
> > composition of EEG and ECoG.
> >
> > Yes, that fits my following understanding.
> > When we consider the recording EEG at scalp, the distance from the
> > electrode to the pole center of a cortical dipole is about 10-15 mm
> > according to the Electric Fields of the brain. Since the pole distance is
> > about 3-4 mm, the cortical source can be approximated as a dipole. If a
> > distance from the measurement point is 3-4 times larger than the pole
> > distance, the dipole approximation works well. This is explained in the
> > Electric Fields of the Brain (Nunez and Srinivasan, 2006) p.109.
> >
> >> Therefore, the amplitude of the (summarized) recorded activity depends
> on
> >
> > the distance between the place of generation and the point of activity
> > registration.
> >
> > For the sub-cortical recording, see Jorge's following paper. It's a
> > good one. You can find empirical evidence for your (theoretical) claim. I
> > visited their lab in early 2010's.
> >
> > Pitfalls in the dipolar model for the neocortical EEG sources.
> > Riera JJ, Ogawa T, Goto T, Sumiyoshi A, Nonaka H, Evans A, Miyakawa H,
> > Kawashima R
> > J Neurophysiol. 2012 Aug; 108(4) 956-975
> > DOI: 10.1152/jn.00098.2011, PMID: 22539822
> >
> >> I would like to offer an explanation by the mutual synchronization of
> the
> >
> > sources of oscillations in the presence of a connection between them.
> >
> > Thank you for sharing your paper.
> > In your 2019 paper I noticed the thalamus is not modeled, but I also
> > thought that thalamus may be still there as an input and modulating
> factor.
> > Do you have any comments on that?
> >
> > Makoto
> >
>